Spatial heterogeneity in soil resources, such as essential nutrients, exerts a profound influence in creating and maintaining high species diversity in plant communities, including the herbaceous layer of forest ecosystems. Indeed, there is wide variation among plant species in requirements for nitrogen (N), with available N typically exhibiting notable heterogeneity in forest soils. Whereas most plant species are competitive at low levels of N (N-efficient species), far fewer require high N availability (nitrophilic species). Thus, a patchy distribution of available N would allow coexistence of both groups of species, enhancing species diversity. Conversely, high availability of N, coupled with low spatial heterogeneity, would result in lower diversity. An early observation in an ongoing whole-watershed N experiment at Fernow Experimental Forest (FEF), West Virginia, was that not only did the N treatment predictably increase extractable soil NO3- and soil solution NO3-, but that the treatment also substantially decreased the spatial heterogeneity of these variables in the 35 ha treatment watershed. This observation was developed into a mechanistic hypothesis to explain the response of forest herb biodiversity to excess N—the nitrogen homogeneity hypothesis.
Results/Conclusions
This hypothesis comprises several facets: (1) the competitive interplay between nitrophilic species and N-efficient species, (2) the maintenance of high species diversity in plant communities with spatial heterogeneity in soil N, and (3) the observation that experimental additions of N can decrease spatial heterogeneity (i.e., increase homogeneity) of available N. Since its initial articulation in 2006, the N homogeneity hypothesis has been discussed in review papers and explicitly/implicitly addressed in several published studies, providing broad support. However, these initial predictions were made before they could be tested adequately at FEF. Now, 25 yr after initiation of N treatments to WS3, we can more fully address the prediction that excess N deposition will increase homogeneity of N availability, and do so in a way that increases homogeneity of the herbaceous layer community. Measured net nitrification in soil at 100 locations in each of two watersheds revealed a range of 0.07 to 118 mg N m−2 d−1 (coefficient of variation = 51%) for the untreated watershed and moderate to high rates (~40 mg N m−2 d−1; CV = 33%) for the N-treated watershed, thus showing greater spatial homogeneity in available N. From 1991 to 2014, herb layer homogeneity decreased on the reference watershed by ~25%, whereas homogeneity increased on the N-treated watershed by >10%, supporting the predictions of the hypothesis.